effect of organosulphur, organonitrogen and organooxygen compounds on the hydrodechlorination of...

Effect of organosulphur, organonitrogen and organooxygen compounds

on the hydrodechlorination of tetrachloroethylene over Pd/Al2O3

Elena Lopez, Fernando V. Dıez, Salvador Ordonez *

Department of Chemical and Environmental Engineering, University of Oviedo, Julian Claverıa-8, 33006 Oviedo, Spain

Received 20 November 2007; received in revised form 25 January 2008; accepted 4 February 2008

Available online 9 February 2008

Abstract

The performance of a 0.5 wt.% Pd on alumina catalyst used for the hydrodechlorination of tetrachloroethylene, in presence of model

organosulphur (thiophene and butanethiol) organonitrogen (quinoline and n-butylamine) and organooxygen (tetrahydrofurane, isobutanol)

compounds, was studied in this work. Experiments were carried out in a continuous fixed bed reactor (space time of 1.8 min g/mmol of TTCE)

at a pressure of 0.5 MPa and a temperature range of 200–300 8C. Concentrations of heteroatomic molecules in the range 0.5–5% were used in this

study.

Organosulphur compounds produce strong inhibition on the TTCE hydrodechlorination, decreasing the conversion and increasing the

selectivity for partially dechlorinated compounds (trichloroethylene). However, this effect is shown to be highly reversible, the catalyst almost

recovering its initial activity when the sulphur source is removed from the feed. Organonitrogen compounds cause a fast and fatal deactivation of

the catalysts, being this effect completely irreversible. Different regeneration procedures were tested, being the treatment with hydrogen at 400 8Cthe only way to partially recover catalytic activity. Finally, organooxygen compounds hardly affect catalyst performance.

# 2008 Elsevier B.V. All rights reserved.

Keywords: Tetrachloroethylene; Hydrodechlorination; Pd catalyst; Organosulphur compounds; Organonitrogen compounds; Organooxygen compounds;

Hydrotreating

www.elsevier.com/locate/apcatb

Available online at www.sciencedirect.com

Applied Catalysis B: Environmental 82 (2008) 264–272

1. Introduction

The management of organic wastes from dry-cleaning

facilities and textile industries is a major environmental

problem. In the most of the modern installations, tetrachlor-

oethylene-based solvents are used for textile cleaning. In order

to enhance the properties of the solvent (detergency, viscosity,

etc.), variable amounts of other organic compounds (such as

hydrocarbons, amines, esters) are added to the chlorinated

solvent TTCE [1]. In the modern installations, the used solvent

is distilled, purified TTCE being reused in the process.

However, the bottoms of this process have high concentration

of TTCE (up to 40%), as well as most of the additives of the

fresh solvent and greases coming from the cleaning process.

This waste (viscous liquid or solid, depending on operation

conditions), which is water-insoluble and very soluble in

organic solvents, is considered as a hazardous waste because of

* Corresponding author. Tel.: +34 985103437; fax: +34 985103434.

E-mail address: [email protected] (S. Ordonez).

0926-3373/$ – see front matter # 2008 Elsevier B.V. All rights reserved.

doi:10.1016/j.apcatb.2008.02.001

its high content of TTCE. This chlorinated compound is

considered as a dangerous pollutant because of their both short-

term (skin irritant, causes lung and central nervous system

diseases) and long-term (liver and kidney diseases, potential

carcinogenic effects) effect on human health [2].

Catalytic hydrodechlorination (HDC) could be a safe

alternative treatment for these wastes, as recently stated in

the European Reference Documents (BREF) on Best Available

Techniques for Waste Treatment [3]. Catalytic HDC consists of

reacting the organochlorinated compounds with hydrogen in

the presence of a catalyst, yielding hydrogen chloride, that can

be easily removed, and, if HDC is complete, hydrocarbons,

which can be burned or even recovered as valuable chemicals.

HDC presents relevant environmental advantages, such as the

no formation of harmful by-products as Cl2, fragments of

parent chlorocarbons, COCl2 or other oxidation by-products,

very common in oxidative treatments. This technique was also

proposed for the treatment of gaseous [4] and aqueous streams

[5].

In order to operate at mild conditions, HDC of most

chlorinated compounds requires the presence of a catalyst. The

mailto:[email protected]

http://dx.doi.org/10.1016/j.apcatb.2008.02.001

Table 1

Composition and textural properties of catalyst Engelhard Escat 16

Composition (wt.%) 0.5% Pd/Al2O3

BET specific surface (m2/g) 92

BJH desorption pore volume (cm3/g) 0.46

Average pore diameter (nm) 18

E. Lopez et al. / Applied Catalysis B: Environmental 82 (2008) 264–272 265

most extensively studied catalysts for HDC reaction in organic

medium are hydrotreatment catalysts (typically promoted metal

sulphides).

They operate at relatively severe conditions (typically

10 MPa, T > 350 8C), and are not stable in the presence of HCl

[6]. Precious metals (Pd, Pt, Rh) are more expensive, but are

more active and work at milder conditions (0.1–0.5 MPa,

250 8C). Research carried out by our group on HDC of chloro-

aliphatic compounds over different precious metals and

supports showed that Pd on alumina presents the best

combination of activity and stability [7]. The influence of

process parameters, such as temperature, pressure, hydrogen/

organic compound ratio and nature of solvent, on the

performance of this catalyst was also studied [8,9]. Results

indicate that high hydrogen/organochlorinated compound ratio

favours conversion and catalyst stability, the presence or nature

of the organic solvent (it must taken into account that the

treatment of these wastes involves the dissolution in organic

solvents) has no relevant effect, while temperature has a two-

fold effect: higher temperature leads to higher initial

conversion, but faster catalyst deactivation. Generally, it is

observed that the main deactivation cause is the formation of

coke deposits, promoted by the surface acidity increase caused

by the HCl released in the reaction [8,9]. The presence of strong

inhibition effects when different chlorinated are processed

together was also observed [10,11].

However, when catalytic hydrogenation is applied to real

wastes (as in the case of the wastes from dry-cleaning

facilities), other organic compounds could be present. Many of

these compounds are hydrocarbons, which are reported to no

significantly affect the performance of this process [8]. But

some of these organics can contain other heteroatoms, such as

sulphur, nitrogen or oxygen. The presence of these elements, as

well as their hydrogenation products: H2S, H2O and NH3, can

severely affect catalyst performance. Because of their chemical

similarities, there are several works studying parallel hydro-

dechlorination, hydrodenitrogenation (HDN) and hydrodesul-

phurization (HDS), but most of these works deal with

hydrotreatment catalysts [6,12,13]. So, it is well established

that the presence of organosulphur compounds does not affect

catalyst performance, whereas the H2S released in the reaction

even enhances the hydrodechlorination reaction [6]. The

hydroprocessing of molecules with both chlorine and sulphur

(typical in some warfare agents) was successfully studied by

Jang and Spivey [13], not observing any kind of deactivation.

The studies are much scarcer in the case of noble metal

catalysts. Typically, these catalysts were considered not

suitable in presence of organosulphur compounds. However,

recent works are focused on HDS over Pd catalysts, as they are

considered as appropriate catalysts for ultra-deep HDS [14]. In

these cases, organonitrogen compounds (specially aromatic

ones) are observed to severely inhibit HDS reaction [15]. In the

same way, Xia et al. [16] observe that pyridine cause severe

inhibition in the hydrodechlorination of oxychlorinated

compounds over Pd catalysts, whereas other organooxygen

compounds (such THF) does not markedly affect the catalyst

performance. In a previous work, we observed that TPN

inhibits the TTCE hydrodechlorination, being this effect

modelled considering a Langmuir-Hinselwood mechanism

with competitive adsorption of TPN and TTCE [17]. It was also

observed that this inhibition effect decrease both TTCE

conversion and selectivity towards fully dechlorinated pro-

ducts.

In spite of the practical importance of the effects of these

sulphur-, nitrogen- and oxygen-containing organic compounds

on catalytic hydrodechlorination, there are not, to the best of

our knowledge, systematic studies on the effect of this

compounds in the deactivation of the palladium catalysts used

for TTCE hydrodechlorination. So, in this work, the effect of

the main heteroatoms on the HDC of TTCE is investigated. For

this purpose, the performance of a Pd/Al2O3 catalysts will be

studied in presence of different organosulphur (thiophene and

butanethiol, BTT), organonitrogen (quinoline, QNL and

butylamine, BTA) and organooxygen (tetrahydrofurane, THF

and isobutanol, ISB) compounds.

2. Experimental

2.1. Materials

The chemicals used in this work (TTCE, TPN, BTT, QNL,

BTA, THF, ISB, trichloroethylene (TCE), toluene, methylcy-

clohexane and decahydronaphtalene) were supplied by

Panreac, Fluka, Janssen and Merck, with a minimum purity

of 99%. Hydrogen C-50 was supplied by Air Products with a

minimum purity of 99.999%. The catalyst tested was Engelhard

Escat 16, a commercial g-alumina-supported Pd catalyst, with

composition and textural characteristics given in Table 1. The

catalyst is available as pellets, which were crushed and sieved

to a particle size between 0.250 and 0.355 mm. Inert a-alumina

(Janssen) was used to dilute the catalyst introduced in the

reactor.

2.2. Reaction studies: equipment and experimental

procedure

Reactions were carried out in a fixed-bed reactor consisting

of a 9 mm i.d., 450 mm long stainless steel cylinder, placed

inside a PID-controlled tubular electric furnace and equipped

with five thermocouples at different reactor heights for

temperature monitoring. The catalyst (typically 0.5 g) mixed

with a-alumina was placed in the mid-section of the reactor.

The bottom and top sections were packed with 1 mm glass

spheres, the upper part being used as pre-heating zone. Before

starting the reaction, the catalyst was activated in situ by

passing through the reactor, heated at 350 8C, 0.90 L/min

(measured at standard conditions) of hydrogen at 0.5 MPa for

E. Lopez et al. / Applied Catalysis B: Environmental 82 (2008) 264–272266

6 h. The liquid and gas feeds flowed co-currently downwards

through the reactor. The liquid feed (toluene, 10% (w/w) of

TTCE and S, N or O-containing compound) was pumped by a

Kontron T-414 liquid chromatography pump, working at space

time of 1.8 g min/mmol of TTCE. At reaction conditions, the

liquid feeds were completely vaporized. The reactor gas feed

was pure hydrogen, with flow rate controlled by a Brooks 5850

TR/X mass-flow regulator.

The reaction products were collected in a stainless steel

Teflon-lined cylindrical receiver. The top of the receiver was

connected to a Tescom 26-1723-24 back-pressure regulator,

which maintained the operating pressure (0.5 MPa, selected to

ensure that most reaction products remained in liquid phase).

Liquid samples were collected by emptying the receiver at

selected time intervals. All the elements of the reactor set up

were constructed in Hastelloy-C, resistant to the corrosion

caused by HCl. The set up was fitted with safety features, such

as temperature and pressure controls, and a rupture disk.

2.3. Analysis and catalyst characterization

Reaction products were analysed by gas chromatography in

a Shimadzu GC-17A apparatus equipped with a FID detector

and a HP-1 30 m capillary column, using decahydronaphtalene

as internal standard. The oven was maintained at 30 8C for an

initial period of 15 min, and then heated to 180 8C at 6 8C/min.

Peak assignment was performed by GC-mass spectra and

responses were determined using standard calibration mixtures.

In all the experiments reported in this work, tetrachlorethylene

hydrodechlorination only yields trichloroethylene and ethane.

Fractional conversions and selectivities were calculated from

the concentration of reactants and products in the inlet an outlet

streams. Chlorine mass-balance was periodically checked by

measuring the HCl concentration in the outlet gaseous stream

(previous solution in alkaline water and chloride tritation). The

formation of NH3 and H2S was followed with an electro-

chemical detector (Drager MiniWarm), whereas the formation

of permanent organic gases (such as ethane or propane) was

determined by GC, using a Carbowax column and a TCD

detector.

When the reaction was stopped, the catalyst was cooled by

flowing nitrogen through the system, separated from the inert

alumina by sieving and collected for subsequent characteriza-

tion. Catalysts samples were characterized by several

techniques. Nitrogen adsorption measurements were performed

in a Micromeritics ASAP 2000 apparatus. Values of surface

area and pores volume were calculated using the methods of

Brunauer-Emmett-Teller and Barrett-Joyner-Halenda, respec-

tively. Thermogravimetric studies were carried out in order to

evaluate the carbonaceous deposits on the used catalysts in a

Mettler TA4000-TG50 thermobalance, in a synthetic air

oxidant atmosphere. Temperature-programmed oxidation

(TPO) and reduction (TPR) studies were performed in a

Micromeritics TPD-2900 apparatus, equipped with TCD

detector, and connected to a Gaslab mass spectra analyser.

Studies were carried out by flowing 29 mL/min of 2 vol.% O2/

balance He (TPO) or 10 vol.% H2/balance Ar (TPR) on the

catalyst samples heated at 10 8C/min. Morphological studies

were made by scanning electron microscopy (SEM) in a JSM-

6100 JEOL microscope, equipped with a EDX Link eXL-1000

microanalysis apparatus, and by transmission electron micro-

scopy (TEM) in a JEOL JEM2000EXII microscope operating

between 160 and 180 kV.

3. Results and discussion

3.1. Effect of organosulphur compounds

3.1.1. Reaction studies

The performance of the catalysts in presence of the studied

heteroatomic compounds was studied at a space time of

1.8 min g/mmol of TTCE, 0.5 MPa total pressure and a

temperature range of 200–300 8C. At these reaction conditions,

it was observed in previous works that the catalyst performs

without significant deactivation for more than 100 h for the

hydrodechlorination of TTCE in absence of the studied

compounds [9]. In the same work, it was observed that even

the effect of externally added HCl on catalyst stability is not

significant up to concentration of HCl of 5%.

The first set of experiments was carried out feeding the

reactor with a liquid feed consisting of 0.5 mL/min of 10 wt.%

TTCE dissolved in toluene, adding 0.5 wt.% of organosulphur

(TPN or BTT) at selected time intervals. The reactor gas feed

was 2.3 L/min (STP) pure hydrogen that corresponds to 50

times the stoichiometric amount (calculated supposing that all

TTCE reacts to ethane, all BTT and TPN react to butane and 5%

of toluene reacts to methylcyclohexane). Experiments were

carried out at 5 bar and 250 8C. The liquid feed contained only

TTCE in toluene during the first 50 h; then, sulphur compound

was continuously added during the next 28 h, or up to total

deactivation. Finally, the feed was changed again to TTCE in

toluene free of organosulphur compound.

In all the cases, the reaction products detected were TCE,

ethane, butane and methylcyclohexane (by GC) and hydrogen

chloride and hydrogen sulphide. The reactions taking place are

the following:

C2Cl4þH2 ! C2Cl3H þ HCl (A)

C2Cl3H þ 4H2 ! C2H6þ 3HCl (B)

C4H10S þ H2 ! C4H10þH2S (C)

C6H5CH3þ 3H2 ! C6H11CH3 (D)

In previous works, we have observed that the reactivity of

TTCE and TCE is very similar, whereas dichloroethylenes and

vinyl-chloride are very reactive [10]. This is justified in the

literature considering a reaction mechanism consisting of

successive hydrogenation–dehydrochlorination steps. The first

step is normally the controlling one, being more favoured as the

number of chlorine atoms joined to the double bond decreases

[18]. This is the cause why other chlorinated compounds are not

detected in quantifiable amounts.

Fig. 1 shows the evolution of TTCE conversion and TCE

selectivity with time on stream, when the effect of BTT on

Fig. 1. Evolution of TTCE conversion (^) and TCE selectivity (*) with time

on stream (TOS) for the hydrodechlorination of 10 wt.% TTCE dissolved in

toluene at 5 bar and 250 8C. BTT (0.5%, w/w) was added during the period 50–

78 h.



toluene at 5 bar and 200 8C. TPN (0.5%, w/w) was added during the period 50–

115 h.


catalyst performance was studied. TTCE conversion decreases

sharply to nearly 0 when BTT is added to the reaction feed.

When BTT is removed from the feed, TTCE conversion

increases gradually, although the recovery in TTCE conversion

is not complete. The evolution of TCE selectivity is the

opposite: TCE selectivity increases in presence of BTT and

decreases again in absence of BTT. After 170 h on stream, the

catalyst is still active for TTCE hydrodechlorination, although

TTCE conversion is significantly lower (about 20%) than

before the BTT addition, and TCE is formed in appreciable

amounts.

The effect of the TPN on TTCE hydrodechlorination is

shown in Fig. 2. In general terms, the behaviour is similar to the

one observed for the same reaction in presence of BTT. TTCE

conversion drop in presence of TPN was less pronounced, and

the recovery of TTCE conversion and TCE selectivity when

TPN was removed from the reaction feed was almost complete.

This result (showing that the effect of TPN is essentially

inhibition rather than deactivation) is in a good agreement with



toluene at 5 bar and 250 8C. TPN (0.5%, w/w) was added during the period 50–

110 h.

literature findings that model the kinetics of the hydrogenation

of TTCE-TPN mixtures using Langmuir-Hinshelwood kinetics

considering competitive adsorption of both organic compounds

[17].

In order to evaluate the effect of the temperature on the

poisoning effect, the experiment was repeated at 200 8C.

Results, depicted in Fig. 3, show that the decrease in the TTCE

conversion and the increase in the TCE selectivity are more

marked as the temperature decreases. This fact suggests that the

adsorption of the TPN plays a key role in the catalyst poisoning.

It should be noted that adsorption equilibriums are exothermic

and hence favoured as temperature decreases.

Fig. 4 shows the evolution of BTT and TPN conversions with

time on stream during the period in which they are fed to the

reactor. It can be observed that initial conversion is notably

higher for BTT, and that both conversions decrease gradually

with time. This result is in a good agreement with the reported

trends for hydrodesulphurization, being traditionally accepted

that terminal sulphur atoms are easier to remove [19]. The

Fig. 4. Evolution of BTT (^) and TPN conversions with time on stream (TOS)

during the period at which the sulphur-containing compound was added to the

reaction feed. TPN conversion data is reported at 200 8C (~) and 250 8C (*).

Table 2

Summary of the main parameters of the deactivated catalysts studied in this

work: fresh catalyst (A), catalyst deactivated at 200 8C in absence of heteroatom

compound (B), catalyst deactivated at 200 8C in presence of TPN (C), catalyst

deactivated at 250 8C in presence of TPN (D) and catalyst deactivated at 250 8Cin presence of QNL (E)

Catalyst

A B C D E

Specific surface area (BET) 92 86 85 85 60

Average crystallite diameter (TEM) 8.2 12.3 11.9 12.2 a

Weight loss (200–800 8C, TG) 0 4.1 2.1 3.1 8.1

a Not measurable.

Fig. 5. Evolution of the CO2 (A) and SO2 (B) releases with temperature during

the TPO analysis of the catalysts aged in absence of TPN (200 8C) and in

presence of a 0.5% of TPN (at 200 and 250 8C).


higher conversion attained by BTT, resulting in higher

hydrogen sulphide concentration in the reaction mixture, can

explain the stronger inhibition observed for BTT. In the same

way, results suggest that H2S plays a key role in the irreversible

deactivation.

3.1.2. Catalyst characterization

In order to elucidate the effect of the presence of

organosulphur compounds on the catalyst, different catalyst

samples were characterized. For this purpose, three different

deactivation experiments were carried out for 23 h reaction

time: working with the conventional TTCE-toluene feed

(200 8C) and adding 0.5% of TPN (at 200 and 250 8C).

Measured BET surface areas (summarised in Table 2) were

92 m2/g for the fresh catalyst, 86 m2/g for the catalyst used for

TTCE hydrodechlorination with no TPN and 85 m2/g for the

catalyst in presence of TPN (at both temperatures). Hence, the

presence of TPN has no relevant effect on the BET surface area.

SEM observations at 10,000 magnifications showed no

significant morphological changes, between all the aged

catalysts.

Pd crystallite size distributions were determined by TEM,

measuring the size of more than 80 crystallites for samples of

catalyst, fresh, reacted with TTCE at 200 8C for 23 h, and

reacted at the same conditions in presence of TPN. The

resulting average crystallite size, calculated according to the

procedure outlined in reference [9] is 8.2 nm for the fresh

catalyst, 12.3 nm for the catalyst reacted in absence of TPN,

and 11.9 and 12.2 nm for the catalyst reacted with TPN at 200

and 250 8C, respectively (Table 2). These results indicate that

the presence of sulphur compounds does not markedly modify

the dispersion of the active phase. So, although in the literature

is reported that sulphur enhances metal particle sintering at

relatively low temperatures [20], this behaviour has not been

observed in this work.

The amount of carbonaceous deposits on the catalysts was

measured by thermogravimetry, according to the procedure

outlined in the literatures [21,22]. The values of weight loss

(between 200 and 800 8C) and main combustion peak are 4.1%

and 404 8C for the catalyst reacted in absence of TPN, 2.1% and

390 8C for the catalyst reacted in presence of TPN at 200 8Cand 3.1% and 405 8C for the catalyst reacted in presence of

TPN at 250 8C. It should be noted that there are not correlation

between the BET surface area and the coke content. This fact

suggests that coke is deposited in thin layer over the support,

not affecting the pore volume. This behaviour is commonly

found both a mesoporous support is used and coking is not the

main deactivation cause [8].

These three samples were also analysed by TPO-MS, in

order to determine the nature of the weight losses. TPO-MS

profiles are shown in Fig. 5A (CO2 release signal) and Fig. 5B

(SO2 release signal). In the case of HCl release, it follows a very

similar trend for all the studied catalysts, being released at

temperatures higher than 600 8C.

Results, which are in good agreement with the TG

parameters, shows that the amount of coke of the aged

catalysts is markedly lower in the case of TPN-treated catalysts.

It is assumed that the combustion temperature of carbonaceous

deposits depends on its chemical structure and the strength of

the interaction between the deposit and the catalyst [7,22].

From this point of view, there are not relevant differences in this

temperature, revealing that the sulphur compound does not

modify the coke properties. However, important differences are

observed in the amount of coke of the material. In general, it is

observed that TPN-treated catalysts bear lower amounts of

coke, decreasing this amount with the reaction temperature.

Fig. 6. Evolution of TTCE conversion (^), TCE selectivity (~) and QNL

conversion (*) with time on stream (TOS) for the hydrodechlorination of

10 wt.% TTCE dissolved in toluene at 5 bar and 250 8C. QNL (0.5%) was added

during the period 0–50 h.

Fig. 7. Evolution of TTCE conversion (^), TCE selectivity (~) and BTA

conversion (*) with time on stream (TOS) for the hydrodechlorination of

10 wt.% TTCE dissolved in toluene at 5 bar and 250 8C. BTA was added during

the period 0–25 h.


The last point is a good agreement with the behaviour observed

during the HDC of tetrachloroethylene [8,9], whereas the

former one suggests that the presence of sulphur compounds

also inhibit the formation of coke. This fact is explained

considering that coke formation is catalysed by both metal

particles and alumina acid sites [9,23], TPN being adsorbed on

both types of sites.

Concerning SO2 release, it is observed only in the case of the

catalyst poisoned at the lowest temperature, being not associated

to CO2 emission. This result suggests that this release

corresponds to the decomposition of labile metal sulphides.

These sulphides are considered in the literature as an important

deactivation cause for hydrogenation reactions [24,25]. The

formation of these sulphides involves the adsorption of the H2S

formed during the TPN hydrogenation, and the reaction of the

H2S with the Pd. The first one is an exothermic process (favoured

as temperature decrease), whereas the second is endothermic

(favoured at increasing temperatures). Our results suggest that

the first one is the controlling step.

From the results obtained, it could be concluded that the

adsorption of organosulphur compounds leads to strong

inhibition effects (reversible) whereas the formation of Pd

sulphides could lead to irreversible deactivation.

3.2. Effect of organonitrogen compounds

3.2.1. Reaction studies

The effect of the presence of organonitrogen compounds on

the catalytic hydrodechlorination of TCE has been studied by

adding 0.5 wt.% of quinoline or butylamine (model for

aromatic and aliphatic compounds, respectively) to the reaction

feed (10 wt.% TCE in toluene). The experimental procedure

and conditions were similar to those reported in the previous

section. The reaction products detected were ethane, hydrogen

chloride, TCE and traces of 1,2,-dichloroethene (from TTCE)

and methylcyclohexane from toluene. Concerning to the

nitrogenated compounds, BTA hydrogenation products were

butane and ammonia, whereas the hydrodenitrogenation of

QNA yields ammonia and a complex mixture of products

including propylbenzene, o-propylaniline, tetrahydroquino-

leine, 1,2,3,4,-tetrahydroquinoline, 1-ethyl-2,3-dimethylben-

zene and 2-amine-5-chloropyridine. In the case of aliphatic

amines, it is well accepted that the reaction follows a direct C–

N bond hydrogenation, yielding the main reaction products

[26]. However, in the case of aromatic organonitrogen

compounds reaction mechanisms are more complex, since

the direct breakage of C–N bond is not possible in these

structures. So, different mechanisms are proposed in the

literature, including dearomatization–elimination and substitu-

tion steps [26,27]. It is also reported that the presence of other

inorganic gases, such as H2S can participate in the reaction

pathways through amine substitutions [28]. Although, a

detailed study of these reaction pathways is out of the scope

of this work, the presence of HCl could largely modify the

reaction pathways.

The evolution with time on stream of TTCE conversion,

TCE selectivity and QNL or BTA conversion are represented in

Figs. 6 and 7, respectively. The presence of either QNL or BTA

leads to a sharp decrease in TTCE conversion and an increase in

TCE selectivity. Thus, the catalyst is completely deactivated

after 10 h on stream in presence of organonitrogen compound.

Conversions of nitrogen-containing compounds, initially very

high, also drop. When QNL or BTA are removed from the

reaction feed, conversions remain near zero, without appreci-

able deactivation. In both experiments, a gradual increase in the

pressure drop through the reactor was observed.

3.2.2. Catalyst characterization

The catalyst recovered after the reactions showed the presence

of solid deposits, visible to the naked eye, which are supposed to

cause the aforementioned gradual increase in pressure drop. The

deposits corresponding to the experiment with QNL were

dissolved in 250 mL water, and the resulting solution analysed by

the Nessler method, resulting in a concentration of NH4+ in the

solution of 210 ppm. The solid deposits were also evident by

SEM, in the form of approximately 300 nm long particles, and

Fig. 8. Ammonia release profile during the TPR (A) and CO2, HCl and NO2

release profiles during the TPO (B) of the catalyst deactivated in presence of

QNL (250 8C).



toluene in presence of 0.5% of QNL at 5 bar and 250 8C. Catalyst treated with

hydrogen at times marked with dotted lines.


produced an important decrease in surface area and pore volume

(BET surface area of 60 m2/g).

Deactivated catalysts were analysed by TPR (Fig. 8A).

Recorded MS patterns show a sharp peak of ammonia release,

being the maximum of this peak at 327 8C. This peak suggests

the presence of strongly chemisorbed nitrogen compounds,

which would be hydrogenated at these temperatures. On the

other hand, TPO analyses of the deactivated catalysts show that

the combustion of the carbonaceous deposits takes place at

lower temperatures (main combustion peak at 360 8C for the

catalyst poisoned with quinoleine, Fig. 8B), whereas very small

amounts of NO2 are released at the same time, suggesting the

presence of small amounts of nitrogen in these deposits.

These results suggest the presence of two concomitant

deactivation causes: inhibition caused by the strong adsorption

of nitrogen compounds and fouling caused by the formation of

ammonium or alkylammonium chlorides. The first effect is

widely reported in the literature, organonitrogen or even

ammonia are considered as strong inhibitors in different

hydrogenation reactions because of its marked Lewis base

character [12,29–31]. However, to the best of our knowledge,

this fouling behaviour has not been reported yet.

3.2.3. Catalyst regeneration

Three different procedures were tested in order to regenerate

the catalyst deactivated in presence of QNL. In two of these

cases, the reactor was emptied, the solid content recovered,

dried overnight at 80 8C, and the catalyst particles separated

from the inert material by sieving.

The first regeneration method, aimed to separate the ionic

solids deposited on the catalyst surface, consists of two

successive washings (during 6 h)–filtering–drying sequences

using acetone (in order to remove organic materials) and

boiling water (in order to remove ionic solids). After this

treatment, the catalyst was filtered, dried overnight at 100 8Cand introduced again in the reactor. The catalyst treated in this

way showed no appreciable recovery of the catalytic activity.

The second method was devoted to solve the solid organic

deposits from the catalyst. It consisted of washing in a Soxhlet

apparatus with toluene and cyclohexene. This method has been

shown to be effective for removing carbonaceous deposits from

hydrogenation catalysts used in reactions with high coke

deposition rates, such as polyaromatics hydrotreating [32]. The

catalyst was then filtered, dried overnight at 110 8C, and

introduced in the reactor. This method did not lead to a

significant recovery of the catalytic activity.

TPR results previously commented, which showed a strong

ammonia peak at 327 8C, suggested another possible method

for catalyst regeneration through the hydrogenation of the

nitrogen compounds. The method consisted of interrupting

periodically the reaction feed during the reaction, and treating

the catalyst in situ with hydrogen above 327 8C (in this case at

400 8C for 15 min, every 8 h). Results obtained are shown in

Fig. 9. This method is partially effective in regenerating the

catalyst, although TTCE conversion decreased continuously,

TCE selectivity increased, and pressure drop through the

reactor increased gradually, so that the reaction had to be

stopped after approximately 30 h. The partial character of this

regeneration can be explained considering that the hydrogena-

tion step lead to only a partial gasification of the deposits, rather

Fig. 11. Evolution of oxygenated compounds conversion with time on stream

(TOS). See codes in Fig. 10.


that the sintering of active phase during the thermal treatment

(catalysts was only 45 min at 400 8C).

From a practical point of view, the usefulness of this method

regeneration method is rather limited, because, although the

bed blocking by the deposition of solids could be diminished

using bigger catalyst particles, or other reactor configurations,

the reactivation method is not able to prevent the pronounced

drop in the yield of total hydrodechlorination products.

3.3. Effect of organooxygen compounds

The effect of the presence of organooxygen compounds on

the catalytic hydrodechlorination of TTCE has been studied by

adding tetrahydrofurane (model for aromatic compounds) and

isobutanol (model for aliphatic compounds) to the reaction

feed. The experimental procedure and conditions were similar

to previous experiments, working at a temperature of 250 8C.

The main reaction products detected were ethane, hydrogen

chloride and TCE; methylcyclohexane (from toluene); water,

butane and traces of ethanol, 1-chloro-2-methybutane, cyclo-

butane, 2,2-dimethycyclobutanone and 2,3-butanediol (from

THF). The conversion of organooxygen compound was very

high in most of the experiments (especially in the case of ISB),

producing appreciable amounts of water. Thus, these experi-

ments also provide valuable information about the effect of

water on the hydrodechlorination reaction. Experiments

carried-out adding to the feed 0.5 or 5 wt.% of each

organooxygen compounds are depicted in Fig. 10. Conversion

of the oxygenated compounds (Fig. 11) also remained constant

with time on stream. It is observed that lineal compound is more

reactive than heterocyclic one, in good agreement with the

behaviour reported in the literature for the hydrogenation of

these compounds [33].

Unlike the rest of the heteroatomic compounds, oxygenated

compounds lead to neither deactivation nor inhibition in the

hydrodechlorination reaction. In the case of the highest

concentration of isobutanol, its presence even enhances the

performance of the hydrodechlorination reaction. Considering

Fig. 10. Evolution of TTCE conversion with time on stream (TOS) at 250 8C,

working with different concentration of oxygenated compounds: without

oxygenated compounds (^), 0.5% of THF (*), 5% of THF (~), 0.5% of

ISB (�) and 5% of ISB (&).

that the deactivation of Pd catalysts is mainly caused by the

formation of coke deposits and the presence of HCl increases

the trend of the catalysts for bearing these deposits [8,9],

obtained results suggest that the presence of water modify the

interaction of the HCl with the support, slightly altering the

formation of active centres for coke formation.

In general, it is considered in the literature referred to

hydrotreatment processes that organooxygen compounds present

lower adsorption intensities than organosulphur and organoni-

trogen compounds, because of the lower basic character (Lewis)

of the oxygen atom compared to sulphur and nitrogen atoms [34].

In the specific case of hydrodechlorination reactions, Xia et al.

[16] found that oxygenated compounds (THF, ethanol,

isopropanol), when used as solvents for hydrodechlorination

of 2,4,40-trichloro-20-hydroxidiphenileter on Pd/active carbon,

could act as inhibitors, especially THF. However, they carry out

the reactions at lower temperatures, in liquid phase and using the

oxygenated compound as solvent.

In good agreement with the reaction results, the character-

ization of the catalysts (BET, TEM, TG) used for the

hydrodechlorination of TTCE in presence of organooxygen

compounds provides results similar to those obtained for

catalysts used for TTCE hydrodechlorination in absence of

heteroatomic compounds.

4. Conclusions

Sulphur-, nitrogen- or oxygen-containing organic com-

pounds studied produced very different effect on the TTCE

hydrodechlorination. The nitrogenated compounds (QNL and

BTA) produced fast, strong and in practice irreversible catalyst

deactivation, due to the formation of solid deposits that also

caused blocking of the catalytic bed. The sulphured compounds

(TPN and BTT) produced strong, but partially reversible

inhibition in the reaction (caused by adsorption of TPN, BTT

and hydrogen sulphide), produced in the reaction. The

oxygenated compounds (THF, ISB), and the water formed

during heir hydrodeoxygenation, presented no appreciable

effect on the reaction in the conditions tested. These results are

very relevant for the practical implementation of hydrode-

chlorination processes, i.e. they indicate that it is very important


to ensure the absence of nitrogen-containing compounds in the

reaction feed.

Acknowledgements

This research was financed by the Spanish Ministry of

Science and Technology (PPQ2000-0664 and CTQ2005-

09105-C04-4). E. Lopez was supported by the Spanish

Ministry of Education and Culture (AP2000-4442). Catalyst

samples were kindly supplied by Engelhard-Rome.

References

[1] Dow Chemical, Dry Cleaning: A Basic Handbook, Midland, MI, 1984.

[2] Toxicological Profile for Trichloroethylene and Tetrachloroethylene,

Agency for Toxic Substances and Disease Registry (ATSDR), Department

of Health and Human Services, Public Health Service, Atlanta, GA, 1997.

[3] European Commission, Reference Document on Best Available Techni-

ques for Waste Treatment, Seville, 2005 (available at http://eippcb.jrc.es/

pages/FActivities.htm).

[4] Z.M. de Pedro, L.M. Gomez-Sainero, E. Gonzalez-Serrano, J.J. Rodrı-

guez, Ind. Eng. Chem. Res. 45 (2006) 7760.

[5] C. Schuth, M. Reinhard, Appl. Catal. B 18 (1998) 215.

[6] S. Ordonez, H. Sastre, F.V. Dıez, Appl. Catal. B 29 (2001) 263.

[7] S. Ordonez, F.V. Dıez, H. Sastre, Thermochim. Acta 379 (2001) 25.


[9] E. Lopez, S. Ordonez, F.V. Dıez, Appl. Catal. B 62 (2006) 57.

[10] S. Ordonez, F.V. Dıez, H. Sastre, Ind. Eng. Chem. Res. 41 (2002) 505.

[11] E. Lopez, S. Ordonez, H. Sastre, F.V. Dıez, J. Hazard. Mater. 97 (2003) 281.

[12] F. Gioia, F. Murena, J. Hazard. Mater. 57 (1998) 177.

[13] B.W.L. Jang, J.J. Spivey, Catal. Today 55 (2000) 3.

[14] A.N. Rothlisberger, R. Prins, J. Catal. 242 (2006) 207.

[15] R. Prins, M. Ergorova, A. Rothlisberger, Y. Zhao, N. Sivasankar, P.

Kukula, Catal. Today 111 (2006) 84.

[16] C. Xia, J. Xu, W. Wu, Q. Luo, J. Chen, Q. Zhang, X. Liang, Appl. Catal. B

45 (2003) 281.

[17] E. Lopez, S. Ordonez, F.V. Dıez, Catal. Today 84 (2003) 121.


[19] S.J. Liaw, A. Raje, K.V.R. Cari, B.H. Davis, Appl. Catal. A 123 (1995)

255.

[20] K. Thomas, C. Binet, T. Chevrau, D. Cornet, J.P. Wilson, J. Catal. 212

(2002) 63.

[21] P. Dejaife, A. Auroux, P.C. Gravelle, J.C. Vedrine, Z. Gabelica, E.G.

Derouane, J. Catal. 70 (1981) 123.

[22] P. Marecot, A. Akhachane, C. Micheaud, J. Barbier, Appl. Catal. A 169

(1998) 189.

[23] C. Menini, C. Park, E.-J. Shin, G. Tavoularis, M.A. Keane, Catal. Today 62

(2000) 355.

[24] E.J.M. Hensen, H.J.A. Brans, G.M.H.J. Lardinois, V.H.J. de Beer, J.A.R.

van Veen, R.A. van Santen, J. Catal. 192 (2002) 98.

[25] A.M. Venezia, V. LaParola, V. Nicolı, G. Deganello, J. Catal. 212

(2002) 56.

[26] R. Prins, Adv. Catal. 46 (2001) 399-.

[27] S. Kozai, H. Kabashima, H. Hattori, Fuel 79 (2000) 305.

[28] R. Prins, Y. Zhao, N. Sivasankar, P. Kukula, J. Catal. 234 (2005) 509.

[29] C.N. Satterfield, M. Modell, J.F. Mayer, AIChE J. 21 (1975) 1100.

[30] G.C. Laredo, E. Altamirano, J.A. de los Reyes, Appl. Catal. A 243

(2003) 207.

[31] G. Yuan, M.A. Keane, J. Catal. 225 (2004) 510.

[32] J.J. Alvarez, S. Ordonez, R. Rosal, H. Sastre, F.V. Dıez, Appl. Catal. A 180

(1999) 399.

[33] P. Grange, E. Laurent, R. Maggi, A. Centeno, B. Delmon, Catal. Today 29

(1996) 297.

[34] M.J. Girgis, B.C. Gates, Ind. Eng. Chem. Res. 30 (1991) 2021.

http://eippcb.jrc.es/pages/FActivities.htm

http://eippcb.jrc.es/pages/FActivities.htm